Team_1
Introduction
The analysis within this report is intended to inform and better equip the marketing manager of an alcohol distributor entering Iowa’s market. An analysis of per capita sales across regions of Iowa will demonstrate where consumption of alcohol is highest, differentiated between different alcohol categories. This will allow the marketing manager to better allocate resources and efforts to regions where they can be most effective in increasing sales.
Data Summary
Demographic Analysis:
The data analyzed in this report contains sales data from 2012 to 2016, measuring both dollar and volume sales, and demographic data for all regions of Iowa. The data is also segmented into three layers: zipcode level, county level, and city level. The focus of the analysis in this report is on per capita consumption and sales, so the first step of our analysis is to merge the demographic data with the sales data. This is done in R as opposed to Tableau, as R allows greater control of how the datasets are merged and the creation of the needed per capita variables.
Race/Ethnicity and Income Analysis
General Sales analysis of County Level Across Categories
General Sales analysis of City Level Across Categories
Across City and County level, the sales data projected similar pattern as Whisky, Vodka and Rum dominated the most part of the sales, followed by Misc, Tequila, Brandy, Schnapps, Gin, etc. The rankings across the top 10 counties with most sales in dollars and most sales in volume stayed constant, with Polk, Linn and Scott county taking the first three across dollar sale and volume sale.
City sales saw identical ranking in popularity of product categories. However, there is change of ranking between sale in dollar and sale in volume. Cities including Waterloo, Iowa city and West Des Moines swtiched places, the reason of this swtich in places could due to the fact that Vodka’s retail prices are lower but sold in greater volume, hence causing the change in ranking, besides this, the pattern of the city level analysis stayed constant with the county level.
Data Analytics
Per Capita Sales Distribution Across Geographies
Research Question 2:
Top 10 geographies for per capita consumption across every liquor category (With R code)
County Level
Calculating total population and sale volume by County and category
summarize(total_population = sum(population),
total_sale_volume = sum(sale.volume))
Calculating per capita consumption by dividing total sale volume by total population
per_capita_data <- summarized_data %>%
mutate(per_capita_consumption = total_sale_volume / total_population)
Rank geographies by per capita consumption across all categories
counties_ranked_data <- per_capita_data %>%
group_by(category, county) %>%
summarize(avg_per_capita_consumption = mean(per_capita_consumption)) %>%
arrange(category, desc(avg_per_capita_consumption)) %>%
group_by(category) %>%
slice(1:10)
Top 10 Counties per capita consumption
City level
Calculating total population and sale volume by City and category
summarized_data <- grouped_data_cities %>%
summarize(total_population = sum(population),
total_sale_volume = sum(sale.volume))
Calculating per capita consumption by dividing total sale volume by total population
per_capita_data <- summarized_data %>%
mutate(per_capita_consumption = total_sale_volume / total_population)
Ranking geographies by per capita consumption across all categories
cities_ranked_data <- per_capita_data %>%
group_by(category, city) %>%
summarize(avg_per_capita_consumption = mean(per_capita_consumption)) %>%
arrange(category, desc(avg_per_capita_consumption)) %>%
group_by(category) %>%
slice(1:10)
Top 10 Cities per capita consumption
Zipcode Level
Calculating total population and sale volume by zipcode and category
summarized_data <- grouped_data_zipcodes %>%
summarize(total_population = sum(population),
total_sale_volume = sum(sale.volume))
Calculating per capita consumption by dividing total sale volume by total population
per_capita_data <- summarized_data %>%
mutate(per_capita_consumption = total_sale_volume / total_population)
Ranking geographies by per capita consumption across all categories
zipcodes_ranked_data <- per_capita_data %>%
group_by(category, zipcode) %>%
summarize(avg_per_capita_consumption = mean(per_capita_consumption)) %>%
arrange(category, desc(avg_per_capita_consumption)) %>%
group_by(category) %>%
slice(1:10)
Top 10 Zipcodes per capita consumption
Top 10 geographies per capita dollar sales across all categories:
To determine and visualize the top geographies by per capita dollar sales, we use Tableau instead of R for a few reasons. Firstly, Tableau enables us to visually depict the per capita sales data on a map of Iowa by employing the longitude and latitude measures, which are unique to the program. Secondly, we can combine multiple visualizations into a dashboard that can be collectively filtered, which, as discussed later in the report, is helpful when examining the geographic trends for specific alcohol categories. Lastly, and more generally, Tableau has intuitive and powerful formatting tools, which allow our visualizations to be aesthetically pleasing.
In Tableau, for all three levels of geography, the per capita dollar sales are mapped onto a map of Iowa, with their size and color gradient, indicating how large sales per capita are in that region. This map is combined in a dashboard with a top ten list bar chart, with the visual color breakdown highlighting the makeup of different alcohol categories in that region. Finally, a filter is added to the dashboard to examine the per capita dollar sales of specific alcohol categories and the corresponding top ten regions. The results for all three levels of geography can be seen below
Top 10 Cities:
Top 10 Cities per capita
Top 10 Counties:
Top 10 Counties per capita
Top 10 Zipcodes:
Top 10 Zipcodes per capita
Analysis of the top regions for per capita sales:
The intuitive choice regarding where to market alcohol in Iowa may be to choose areas with the highest population density or total sales, given that these areas have the largest potential customer base. However, as seen from the analysis above, the areas with the highest total dollar sales do not necessarily correlate with the highest per capita sales of alcohol. For example, Des Moines, Cedar Rapids, and Davenport are the three largest cities in Iowa and have the highest total dollar sales. However, they are not in the top ten cities for alcohol sales per capita. High per capita sales indicate a substantial demand for alcohol in a region. This fact makes the regions highlighted in the visualizations above potentially good regions to enter the Iowa market. Furthermore, the analysis can be drilled down to the alcohol category level, and regions with the highest per capita sales in a particular category can be identified and marketed to, specifically for that alcohol type.
A Notable Outlier:
When examining the top cities for every alcohol category, a particular city stands out for sales of distilled spirits. In Mount Vernon, there is a much higher sale of distilled spirits per capita than in any other city in the state, and by some margin. This indicates a high demand for distilled spirits in Mount Vernon, making the region a potential target for marketing our distilled spirit brand. This is demonstrated in the screenshot below:
Mount Vernon - An outlier for Distilled Spirits
Conclusion
<<<<<<< HEAD Q2: The most popular liquor category remained constant throughout the county, city and zipcode level, the geographies with most per capita consumption also aligns with the general sales analysis data, supporting the assumption of the popularity of Whisky, Vodka and Rum. Out of all the geographies, Dickinson County and Bevington city stood out, as they have the highest per capita consumption in their geographical category respectively, while not being in the top 10 of overall sale in dollars and volume. This could mean that the drinking culture in these two places is more profound comparing to other counties and cities in the state of Iowa. ======= When entering a new market, it is strategically important to understand that market comprehensively. To achieve this goal, this report merged demographic and sales data across geographic levels and determined regions with a history of substantial demand for our products. The analysis and visualization created empower management and other decision-makers with the knowledge necessary to allocate resources efficiently and confidently.
The focus of this report is per capita sales and consumption throughout the levels of geography. The distribution of per capita sales is not uniform at the zipcode or city level and is heavily right-skewed, with a few outliers having high per capita sales. At the county level, the distribution level is more uniform but is still skewed right with a left tail. The top geographies by per capita consumption and sales are listed below:
{r echo = FALSE} library(knitr) kable(geography_rank_table.csv)
While the analysis in this report is significant, it is limited to three years and does not enable a large-scale historical analysis of the trends within the region. Expended datasets that include more years could reveal further insights about the region as a whole and about particular areas within Iowa. Additionally, a breakdown of the sales by date, as opposed to total annual sales, would enable an analysis of seasonal dynamics, improving our marketing and resource allocation throughout the year.
82eed71ef0a6568c79e80399fc0678d363d0f37e
Policy Recommendation
The recommendation of this report is to begin entering the Iowa market in areas with high per capita consumption in sales and then, depending on the success within these regions, expand into areas with higher population density and total sales overall. This tiered approach will better ensure a low-risk but profitable breakthrough into the market. This is because, in regions with smaller total sales but high per capita sales, we do not need to invest as much capital and supply when entering the market to meet the demand. However, because the per capita sales are high, we can expect a return on that investment, resulting in a profitable enterprise. Then, once our relationships with our customers and our brand recognition are established, we can expand into the high total sales markets. This way, we reduce our potential risk and provide a roadmap to increase our market share.